boost::flat_map性能测试

作者: NickYang 分类: 技术文章,程序开发 发布时间: 2015-06-30 10:11

文章转自:boost::flat_map and its performance compared to map and unordered_map

have run a benchmark on different data structures very recently at my company so I feel I need to drop a word. It is very complicated to benchmark something correctly.

Benchmarking

On the web we rarely find (if ever) a well engineered benchmark. Until today I only found benchmarks that were done the journalist way (pretty quickly and sweeping dozens of variables under the carpet).

1) You need to consider about cache warming

Most people running benchmarks are afraid of timer discrepancy, therefore they run their stuff thousands of times and take the whole time, they just are careful to take the same thousand of times for every operation, and then consider that comparable.

The truth is, in real world it makes little sense, because your cache will not be warm, and your operation will likely be called just once. Therefore you need to benchmark using RDTSC, and time stuff calling them once only. Intel has made a paper describing how to use RDTSC (using a cpuid instruction to flush the pipeline, and calling it at least 3 times at the beginning of the program to stabilize it).

2) rdtsc accuracy measure

I also recommend doing this:

This is a discrepancy measurer, and it will take the minimum of all measured values, to avoid to get a -10**18 (64 bits first negatives values) from time to time.

Notice the use of intrinsics and not inline assembly. First inline assembly is rarely supported by compilers nowadays, but much worse of all, the compiler creates a full ordering barrier around inline assembly because it cannot static analyze the inside, so this is a problem to benchmark real world stuff, especially when calling stuff just once. So an intrinsic is suited here, because it doesn’t break the compiler free-re-ordering of instructions.

3) parameters

The last problem is people usually test for too few variations of the scenario. A container performance is affected by:

  1. Allocator
  2. size of contained type
  3. cost of implementation of copy operation, assignment operation, move operation, construction operation, of the contained type.
  4. number of elements in the container (size of the problem)
  5. type has trivial 3.-operations
  6. type is POD

Point 1 is important because containers do allocate from time to time, and it matters a lot if they allocate using the CRT “new” or some user defined operation, like pool allocation or freelist or other…

Point 2 is because some containers (say A) will loose time copying stuff around, and the bigger the type the bigger the overhead. The problem is that when comparing to another container B, A may win over B for small types, and loose for larger types.

Point 3 is the same than point 2, except it multiplies the the cost by some weighting factor.

Point 4 is a question of big O mixed with cache issues. Some bad complexities containers can largely outperform low complexity containers for small number of types (like map vs. vector, because their cache locality is good, but map fragments the memory). And then at some crossing point, they will lose, because the contained overall size starts to “leak” to main memory and cause cache misses, that plus the fact that the asymptoptic complexity can start to be felt.

Point 5 is about compilers being able to ellude stuff that are empty or trivial at compile time. This can optimize greatly some operations, because the containers are template, therefore each type will have its own performance profile.

Point 6 same as point 5, PODS can benefit from the fact that copy construction is just a memcpy, and some containers can have a specific implementation for these cases, using partial template specializations, or SFINAE to select algorithms according to traits of T.

About the flat map

Apparently the flat map is a sorted vector wrapper, like Loki AssocVector, but with some supplementary modernizations coming with C++11, exploiting move semantics to accelerate insert and delete of single elements.

This is still an ordered container. Most people usually don’t need to ordering part, therefore the existence of unordered...

Have you considered that maybe you need a flat_unorderedmap ? which would be something like google::sparse_map or something like that. An open address hash map.

The problem of open address hash maps, is that at the time of rehash they have to copy all around to the new extended flat land. When a standard unordered map just have to recreate the hash index, but the allocated data stays where it is. The disadvantage of course is that the memory is fragmented like hell.

The criterion of a rehash in an open address hash map is when the capacity overpasses the size of the bucket vector multiplied by the load factor.

A typical load factor is 0.8 therefore, you need to care about that, if you can pre-size your hash map before filling it, always presize to: intended_filling * (1/0.8) + epsilon this will give you a guarantee of never having to spuriously rehash and recopy everything during filling.

The advantage of closed address maps (std::unordered..) is that you don’t have to care about those parameters.

But the boost flat_map is an ordered vector therefore it will always have a log(N) asymptoptic complexity, which is less good than the open address hash map (amortized constant time). You should consider that as well.

Benchmark results

This is a test involving different maps (with int key and __int64/somestruct as value) and std::vector.

tested types information:

Insertion

EDIT:

Ok, because my previous results included a bug, they actually tested ordered insertion, which exhibited a very fast behavior for the flat maps.
I left those results thereunder because they are interesting.
This is the correct test: enter image description here

enter image description here

I have checked the implementation, there is no such thing as a deferred sort implemented in the flat maps here. Each insertion sorts on the fly, therefore this benchmark exhibits the asymptotic tendencies:

map : N * log(N)
hashmaps : amortized N
vector and flatmaps : N * N

Warning: hereafter the test for std::map and both flat_maps here is buggy and actually testsordered insertion:
random insert of 100 elements without reservation

We can see that ordered insertion, results in back pushing, and is extremely fast. However, from non-charted results of my benchmark, I can also say that this is not near the optimiality of back-insertion in a vector. Which is 3Million cycles, we observe 4.8M here for boost (160% of the optimal).

random insert of 10000 elements without reservation

Random search of 3 elements (clocks renormalized to 1)

in size = 100

rand search within container of 100 elements

in size = 10000

rand search within container of 10000 elements

Iteration

over size 100 (only MediumPod type)

Iteration over 100 medium pods

over size 10000 (only MediumPod type)

Iteration over 10000 medium pods

best regards

如果觉得我的文章对您有用,请随意打赏。您的支持将鼓励我继续创作!

11条评论
  • 李阳博客

    2015 年 7 月 12 日 21:31

    哇,你这个评论的这个怎么搞的呀,这么高大上~

    活跃 Chrome浏览器 Windows 7 x64 Edition
    1. eliteYang

      2015 年 7 月 12 日 21:54

      哈哈……前一阵在开源中国上看到这个插件了,https://git.oschina.net/wangjunfeng/wp-level-useragent

      站长 Chrome浏览器 Windows 7 x64 Edition
      1. 李阳博客

        2015 年 7 月 13 日 08:24

        这个人我之前用过他的主题,以前是射雕天龙哈哈

        活跃 Chrome浏览器 Windows 7 x64 Edition
      2. 李阳博客

        2015 年 7 月 13 日 09:44

        不过,xiu主题安装后不显示,自定义也不显示。很纠结,你遇到过这个问题吗?

        活跃 Chrome浏览器 Windows 7 x64 Edition
        1. eliteYang

          2015 年 7 月 13 日 09:47

          清过缓存吗?

          站长 Chrome浏览器 Windows 7 x64 Edition
          1. 李阳博客

            2015 年 7 月 13 日 09:51

            清理过,代码如下:
            echo 111
            echo jw_level_useragent_output_custom();
            111就能显示,下面的那个函数就不显示好奇怪

            活跃 Chrome浏览器 Windows 7 x64 Edition
            1. eliteYang

              2015 年 7 月 13 日 16:00

              那这个不应该啊,你是不是没装对地方,是放在插件目录的,或者你上传的时候没传完整,换成FTP二进制模式试试

              站长 Chrome浏览器 Windows 7 x64 Edition
              1. 李阳博客

                2015 年 7 月 14 日 12:38

                插件是装对了,采用插件设置,评论之前和评论之后虽然可以显示,但是前台看不到,只能在后台的评论功能看到~应该是我没调用对。

                活跃 Chrome浏览器 Windows 7 x64 Edition
                1. eliteYang

                  2015 年 7 月 14 日 13:29

                  那就不知道是啥问题了

                  站长 Chrome浏览器 Windows 7 x64 Edition
  • 小幻

    2015 年 6 月 30 日 16:48

    英文的,看不懂

    潜水 火狐浏览器 Windows 8.1
    1. eliteYang

      2015 年 7 月 1 日 15:08

      还是比较容易看懂的

      站长 Chrome浏览器 Windows 7 x64 Edition

发表评论

电子邮件地址不会被公开。 必填项已用*标注